Abstract
In this article, we conduct an integrative and rigorous review of theory and research on education in emergencies programs and interventions as international agencies implement them in areas of armed conflict. We ask several questions. How did this subfield emerge and what are the key conceptual frameworks that shape it today? How do education in emergencies programs affect access, learning, and protection in conflict-affected contexts? To answer these questions, we identify the conceptual frameworks and theoretical advances that have occurred since the inception of the field in the mid-1990s. We review the theories that frame the relationship between education and conflict as well as empirical research that tests assumptions that underpin this relationship. Finally, we assess what we know to date about “what works” in education in emergencies based on intervention research. We find that with regard to access, diminished or inequitable access to education drives conflict; conflict reduces boys’ and girls’ access to education differently; and decreased distance to primary school increases enrollment and attendance significantly for boys and even more so for girls. With regard to learning, education content likely contributes to or mitigates conflict, although the mechanisms through which it does so remain underspecified; and peace education programs show promise in changing attitudes and behaviors toward members of those perceived as the “other,” at least in the short term. Finally, providing children living in emergency and postemergency situations with structured, meaningful, and creative activities in a school setting or in informal learning spaces improves their emotional and behavioral well-being.
Nearly a quarter of the world’s population, or about 2 billion people, live in countries affected by conflict or extreme criminal violence (World Bank, 2016). Conflict inhibits development, and poor countries that experience conflict typically have higher poverty rates than stable countries (World Bank, 2011). Similarly, millions are affected each year by climate-related disasters and earthquakes that destroy homes, neighborhoods, and schools. In 2011 and 2012 alone, more than 450 million individuals faced environmental crises (Blankespoor, Dasgupta, Laplante, & Wheeler, 2010; Laframboise & Loko, 2012). In the midst of an emergency, when a state bureaucracy is weak or absent, nonstate actors such as United Nations (UN) agencies and other international or local nongovernmental organizations often support educational interventions. For example, they may support schools in communities where state services have broken down, or provide supplemental educational activities that are intended to protect children from harm and to promote cognitive, emotional, and social development (Sinclair, 2001, p. 23).
Aid organizations refer to these foreign interventions as “education in emergencies” programs. The growing interest in the field of education in emergencies stems, in part, from the realization that millions of children live in crisis-affected regions, where they often lack access to education (United Nations Educational, Scientific and Cultural Organization [UNESCO], 2011). Indeed, approximately 50 million primary and lower secondary age children are out of school in conflict-affected countries (Save the Children, 2013). Similarly, natural disasters have a dramatic effect on education (Asian Disaster Preparedness Center, 2008; Save the Children, 2008; United States Agency for International Development, 2014). There are four times as many natural disasters as in the 1980s (Oxfam International, 2007), and the average number is predicted to increase exponentially in the next 20 years (Webster, Ginnetti, Walker, Coppard, & Kent, 2008).
In addition to these stark needs, multiple forces have shaped the field of education in emergencies. At least since the 1950s, when President Truman launched modern international development aid, this work has included support to education in low- and middle-income countries. Yet in the midst of crises or conflict, support often shifts toward humanitarian aid, which has rarely focused on education. In response, humanitarian aid workers have highlighted the gaps in services that crises create or exacerbate. Pointing to refugee children as well as populations affected by crises, they have argued for providing education aid to further development goals of achieving education for all and to increase children’s protection in these unstable environments. More recently, the field has garnered attention from strong states, such as the United States, in part because of the belief that uneducated masses may contribute to instability—or “fragility”—in weak states or countries emerging from crisis and, conversely, that outside aid to education can stabilize these countries, enhance state building efforts, and promote global security.
These approaches to education in crises have been motivated by and involved aid professionals and aid organizations interested in understanding program outcomes. In recent years, the use of methods such as randomized controlled trials (RCTs) that assess aid interventions have enhanced the tools available to funders to evaluate their aid spending and catalyzed education researchers to expand the ways they conduct research in countries affected by crisis. Indeed, the surge in RCTs—also known as impact evaluations—carried out in low- and middle-income countries has increased the demand among international funders for rigorous empirical research in order to determine “what works” to achieve program outcomes.
Although education in emergencies programs are implemented in areas affected by both violent conflict and natural disaster, this article focuses solely on education in emergencies in areas of armed conflict. To date, both scholars and practitioners have focused more on conflict settings, and as a result, there are insufficient data from natural disaster settings to warrant their inclusion here. The lack of information on natural disaster settings represents a substantial gap in the literature and one that we predict will change soon. We also focus on a view of education in emergencies that is linked closely to the international aid community. Although regional and national governments and local communities also engage in activities to mitigate the impact of conflict on education, there is limited research on these efforts and even less work that analyzes or evaluates these initiatives.
In the following pages, we conduct an integrative and rigorous review of the literature from conflict contexts, identifying theoretical, empirical, and intervention research. The theoretical is important for defining circumstances and hypothesizing about the relationships between conflict and education, whereas the empirical tests these hypotheses. Intervention studies are empirical work that helps show what mitigates the effects of conflict on education and of education on conflict. 1 We organize this material into four sections. First, we provide a brief overview of the field and discuss education in emergencies as a conceptual framework. Second, we explain our criteria for selecting material to review. Third, we present our findings, focusing on the three most common categories of educational outcome: access, learning, and protection. For each of the three categories, we review theories that frame the relationship between education and conflict as well as empirical research that tests assumptions that underpin this relationship. Finally, we assess what works in education in emergencies based on intervention research and conclude with a summary of key findings and future directions.
Education in Emergencies: A Conceptual Framework
As humanitarian action evolved and expanded in the mid to late 1990s, 2 international aid workers took the opportunity to promote education as a key element of humanitarian responses. They called these education initiatives “education in emergencies” programs. The term was compatible with language used to promote aid in countries affected by conflict, with an emphasis on urgent support for education as a humanitarian need and human right. In fact, it was implicitly chosen to solidify this link to humanitarian action and ensure the incorporation of support to education among other forms of relief aid. The term was also compatible with an interest in outcomes related to improving lives of beneficiaries, a central goal of aid work.
This approach to education outcomes gained ground in the midst of a confluence of global events that pressured humanitarian action to expand beyond its traditional activities. Humanitarianism came to include a concern not just with individuals’ basic needs but also with fulfilling their rights to a certain standard of life. International human rights instruments such as the Convention on the Rights of the Child in 1989 and follow-up action meetings held by UN agencies expanded, strengthened, and institutionalized the recognition and protection of children’s rights, including the right to education. These trends coincided with the decline of interstate conflicts, which were replaced by longer lasting intrastate conflicts that displaced or killed greater numbers of civilians, including many children (Paris, 2004). The importance of providing support to education during crises became more evident. At the same time, local populations affected by conflict also requested support for education from aid donors. Educators and other practitioners, in turn, noted the gap in services provided to populations affected by crisis and advocated for increased support to education (Anderson, Martone, Perlman Robinson, Rognerud, & Sullivan-Owomoyela, 2006; Cahill, 2010; Karpinska, 2012).
Educators adopted a number of strategies to include education in the humanitarian response paradigm. First, in 2000, a group of educators, linked mainly to organizations such as the UN agencies and the International Rescue Committee (IRC), formed the Interagency Network for Education in Emergencies (INEE) to support education in countries affected by conflict and disaster. In doing so, they defined education in emergencies as a category separate from development activities in order to incorporate education into traditional humanitarian assistance (Burde, Kapit-Spitalny, Wahl, & Guven, 2011). The term emergencies signaled its urgency and underscored the relevance of education to the humanitarian response paradigm. 3 Second, placing emphasis on delivering education as a service that could be packaged (i.e., United Nations Children’s Emergency Fund’s “School in a Box”) highlighted its compatibility with other forms of packaged emergency aid. Third, stressing service delivery had additional implications. It helped distance education from politics, further deepening its compatibility with core humanitarian principles (Burde, 2014). It also resulted in a rhetorical emphasis on access over quality. 4
Today, despite some continued resistance to education as an element of humanitarian action, aid to education is often included in support to crisis-affected countries. Major bilateral donor agencies now earmark funds for education in emergencies programs and hire staff whose portfolios include tracking and monitoring these programs. Although it is important to point out that funding for education in crises continues to lag behind other sectors (Save the Children, 2013; Winthrop, Ndaruhutse, Dolan, & Adams, 2010). Despite this lag, INEE membership has increased from a handful of founders and affiliates in 2000 to more than 11,500 members in 2015.
This rise of investment in education in emergencies contributed to a growing awareness of the importance of providing support to education to those in need. However, another factor fundamentally challenged the framework in which many humanitarians and policymakers viewed education aid. The September 11 attacks triggered increased attention from Western governments to countries affected by conflict. States considered to be “failed,” “failing,” or “fragile” became targets for Western intervention. These countries were no longer perceived to be a danger only to themselves; many Western governments suddenly perceived them as posing an imminent threat to the world at large and to the Western world in particular (Barnett, 2009). Education was identified as one way to mitigate these dangers. Within departments of state (e.g., the United States, the United Kingdom), emphasis shifted from viewing support to education systems as important for distant children’s well-being to viewing this support as important for preserving the security of the homeland.
The new priority on education in fragile states makes specific assumptions about the relationship between education and state stability (Novelli, 2010; Winthrop & Matsui, 2013). Improving the provision of state services such as education is thought to engage youth, defuse radicalization, and increase state legitimacy in the eyes of its citizens. As a result, education responses are often intertwined with political goals, both in terms of identifying states that meet the definition of fragile and in the content of the intervention. Although this language has been widely taken up by Western governments and their bilateral aid organizations (e.g., AusAid, United States Agency for International Development), it has been criticized both for the perceived stigma attached to it and for its link to Western security interests (e.g., Novelli, 2010; Winthrop & Matsui, 2013). INEE has recently renamed its Working Group on Education and Fragility to Education Policy Working Group, and many humanitarian organizations avoid using the term.
Yet the education in emergencies framework is also imperfect, particularly in its use of the term emergency, which implies a temporary condition and seems ill-suited to describe crises that endure over time (Burde, 2014). The phrases “education in crises” and “education in protracted crises” are also used to soften and clarify the connection to emergency while preserving the sense of urgency intended to garner a response. The precise terminology is most relevant for outsiders delivering aid, as aid programs are typically considered simply education or social welfare programs by their recipients. Here we use the phrase “education in emergencies” because it remains the most common term among international practitioners.
Method
Theoretical, qualitative, and quantitative studies 5 examine both the problems that affect access to education, learning, and protection, and the programs that attempt to address these problems by increasing children’s access to education, learning, and protection in countries affected by conflict. As noted, first, we examine the theoretical relationships between emergencies and education. Second, we review research that tests these assumptions and the evidence that supports them. Thus, we explore the effect of conflict on educational access, learning, and protection, and vice versa. Finally, we review interventions that aim to address these problems. We explore what works in these contexts, reviewing the research assessing the impact of program interventions intended to promote access, learning, and protection.
Search Parameters
We included only literature that focused on primary or secondary education—because education in emergencies tends to focus mainly on these groups—in states experiencing or recovering from conflict. 6 In addition, we focused on literature analyzing aid recipient countries. 7 Because education in emergencies is a new field according to the way we, and most authors, conceptualize it, and because our primary goal is to shed light on recent interventions, we included only work published since 1995. Finally, we focused on peer-reviewed journals, although we also included recent academic books dedicated to education in emergencies, seminal policy pieces (determined by the extent to which subsequent publications reference them), and works published by international agencies that met the inclusion criteria.
We conducted our primary search for academic articles in two phases. First, in order to define the parameters of our categories described above, we conducted a broad review, dividing the search terms by subject area and conducting individual searches on (a) access to and quality of education in conflict-affected countries, (b) child protection (psychosocial interventions), and (c) peace/tolerance/human rights education. Each researcher assessed articles for relevance and methods, and wrote analyses of the articles she found to be appropriate.
Based on these findings, we refined our questions and conducted a new search with a more precise focus on access, learning, and protection. This search listed education and any one of the following terms: armed conflict, emergency, fragility, or humanitarian. We searched Academic Search Complete, EBSCO Host, ERIC, JSTOR, Proquest Central, and PsychINFO. We conducted additional searches on the sites of the Abdul Latif Jameel Poverty Action Lab, the National Bureau of Economic Research, and the World Bank because their researchers conduct impact evaluations and serve as sources of knowledge for the international aid community. We reviewed titles and abstracts of approximately 3,000 articles in fall 2013.
From this batch, we selected those that matched our focus and then reviewed them to ensure they fit the aforementioned criteria. We divided the selected work into theoretical, empirical, and intervention-related categories. The first set of articles or books includes those studying the relationship between education and violent crisis conceptually. We included only research explicitly theorizing the causal mechanisms influencing this relationship, excluding pieces that merely described an effect. The second category includes empirical work on the relationship between education and conflict, with the causal arrows pointing in either direction. Third, we included articles or books evaluating the impact of programs designed to intervene in this relationship. We included only qualitative studies with clear descriptions of the methods used, sample drawn, and logic behind the research design. We included only quantitative studies that explained the design and analytical models and reported clearly on statistical results.
Finally, after receiving reviewers’ comments, we added books and articles in summer 2015. First, we conducted a new search only for books using the terms “education” AND “conflict” and “education” AND “emergencies,” using Google Scholar, JSTOR, and seven EBSCO databases related to education. Second, we searched for work that matched our criteria from well-known scholars (determined by numbers of citations, length of time publishing, level of public profile). Third, we reviewed a similar study that two of the authors had conducted in 2015 as part of a rigorous review for the British aid agency and included relevant work that we may have overlooked from that publication (see Burde, Guven, Kelcey, Lahmann, & Al-Abbadi, 2015). Fourth, we conducted a final pass for key articles that we may have overlooked or that may have emerged since our earlier searches. In total, we included 121 articles in this review, with 6 included in more than one category: 45 theoretical pieces, 49 empirical articles that provide evidence on the relationship between education and armed conflict, and 35 empirical articles that assess interventions and provide evidence on what works.
Results
Historically, the field of education in emergencies viewed education as largely neutral, if not wholly positive, leading almost inevitably to a range of positive outcomes including child protection and well-being, economic development, peace building, and reconstruction (Davies, 2004, 2005; Davies & Talbot, 2008; Kagawa, 2005; Novelli & Cardozo, 2008; Paulson & Rappleye, 2007). More recently, however, scholars and practitioners have acknowledged that the relationship between education and conflict is more complex than previously recognized (Brock, 2011; Burde et al., 2011; Davies, 2004; King, 2011, 2014; Østby & Urdal, 2010; Rappleye, 2011; Shields & Paulson, 2015; Smith, 2007). This shift was influenced both by the events of September 11, as noted above, as well as by a policy paper titled “The Two Faces of Education in Ethnic Conflict” (Bush & Saltarelli, 2000). The paper was among the first to point out that education can both fuel hostility as well as foster social cohesion. Starting with issues relating to access to education below, we examine the current, more nuanced state of the field.
Access to Education
Of all the proposed relationships between education and emergencies, how access to schooling either aggravates or ameliorates social tensions has received the most attention. One reason is that access is relatively easy to measure by looking at enrollment numbers and primary and secondary completion rates, and most countries collect these data. A substantial body of scholarship in political science has examined the causes of conflict, which sometimes includes educational opportunity, or access to education, as a key variable. In addition, school attendance is critical before other interventions can take place. This research provides a basis for conceptualizing how conflict reduces access to education and how lack of, or diminished access to, education drives conflict. In order to assess the impact of programs that address these tensions, it is important to understand what we know about how these interactions work.
Diminished or Inequitable Access to Education Drives Conflict
Scholars have theorized that inequalities and disruptions in the provision of education are important factors that fuel conflict. Indeed, a lack of educational opportunity has been considered a predictor of conflict. Young people may be more likely to join armed movements if there are few educational opportunities that offer alternative pathways and hope for the future (Barakat, Karpinska, & Paulson, 2008; Barakat & Urdal, 2009; Rose & Greeley, 2006). Furthermore, social exclusion that occurs when access to education falls along ethnic or ideological lines—or even the perception that it does so—likely fuels tensions (Colenso, 2005; Davies, 2005, 2011; Kirk, 2011). For instance, in Nepal, inequality in access to and quality of education may have exacerbated conflict. The Maoist insurgency has drawn on educational disparities as part of its campaign to discredit the government (Parker & Standing, 2007; Pherali, 2013). Indeed, schools are among the most widespread and visible of state institutions, so exclusion from them may contribute to a group’s grievances against the state. Conversely, equitable access to education can contribute to peace and stability in part because the provision of mass schooling helps increase the legitimacy of the nation-state (Burde, 2014; Mosselson, Wheaton, & Frisoli, 2009).
The theoretical relationship between out of school youth and conflict described above has been established in several empirical studies and has explored why a link between educational access and conflict may exist. For example, youth in Sierra Leone without access to education were nine times more likely to join violent conflict than those who attended school (Humphreys & Weinstein, 2008). Scholars have found that in states with youth bulges, large populations of (male) youth are more likely to experience conflict if their populations have lower levels of education, although they have also found that relationship is mediated by various other factors, such as national resources, poverty, and regime type (Barakat & Urdal, 2009; Collier & Hoeffler, 2004; Collier, Hoeffler, & Soderbom, 2004; Østby & Urdal, 2010). Income level, economic structures, and geographical region all matter for the extent to which education is related to conflict in these youth bulge countries. In one study, the relationship appeared strongest in low- and middle-income countries and in Sub-Saharan Africa (Barakat & Urdal, 2009).
Education also may contribute to state stability through increasing the legitimacy of the state, spurring economic growth, reducing inequality, and building social cohesion. Higher literacy and secondary school enrollment rates decreased the probability of civil conflict in 160 countries analyzed between 1980 and 1999 (Thyne, 2006). Some scholars have found that educational investment is even more strongly related to a lower rate of civil conflict if it is equally distributed, because this decreases grievances against the state (Østby, Nordås, & Rød, 2009; Thyne, 2006). Importantly, Østby et al. (2009) argued that regional identity influences perceptions of unequal distribution of education more than any other kind of social affiliation because (a) regional belonging is often founded on a shared history and it overlaps with other types of identities (e.g., ethnicity) and (b) regional boundaries often shape the distribution of state welfare and political influence, thereby producing the gaps between the “haves” and the “have-nots.” Regional inequalities can thus reinforce competition for state benefits, fostering aggression (Østby et al., 2009). Further research is necessary to understand if regional analysis better captures the relationship between educational inequality and conflict than country-level studies. However, educational inequality may not always increase conflict. One cross-national quantitative study did not find any significant effect for inequality on the risk of civil conflict (de Soysa & Wagner, 2003, as cited in Østby & Urdal, 2010). However, it is not possible to disentangle the unit of analysis (in this case, the state) from findings; a regional analysis might have shown different effects.
These quantitative studies begin to explain why educational access influences the risk of conflict. Case study and ethnographic research enriches our understanding of these mechanisms, such as how local dynamics inform this relationship and how combatants themselves see education as relating to conflict (Østby & Urdal, 2010). For example, in-depth interviews complement survey research to show that youth join rebel movements in part due to lack of access to education (Brett & Specht, 2004).
Significantly, the ameliorating effect of education on the propensity to participate in hostilities has been called into question in the context of terrorism, one type of violence posited to attract disaffected and undereducated youth, and more recently in the absence of other governance reforms. 8 Indeed, Krueger and Maleckova (2003) found no link between education level and support for terrorism. Hezbollah combatants killed in the late 1980s and early 1990s tended to have greater secondary education attainment than noncombatants, and Israeli Jews involved in terrorism during the 1980s tended to be particularly well educated. Given youth anger toward injustice, a recent report suggests that interventions aiming to cultivate civic education in the absence of governance reforms may stoke violence rather than decrease it (Proctor, 2015). Taking into consideration the presence of political dissatisfaction adds important nuance to this discussion. However, the high education levels of some terrorists may suggest that terrorist groups prefer to select highly educated people, not that the educated are more likely to attempt to join them (Berrebi, 2007; Fair, 2007).
Education also interacts with support for terrorist and militant groups and their activities among the general population. For example, Shafiq and Sinno (2011) found that in some countries higher levels of educational attainment indicate greater support for suicide bombing when education is mediated by political dissatisfaction. Similarly, Afzal (2013) found that better educated men in Pakistan are more likely to support the Pakistani Taliban, yet better educated women are less likely to support terrorist organizations. As Afzal showed, disaggregating results according to gender is a critical area of inquiry that deserves more attention.
In sum, much of the data reviewed above suggest that the benefits of increasing access to education outweigh the potential risks. Governments should seek to provide equitable access to education at all levels and for all groups. The post–September 11 focus on fragile states has underscored the importance of examining the relationship between education and conflict and increased interest in assessing programs designed to mitigate these effects. To understand better how education affects conflict, it is important to assess the role education plays as it interacts with other factors in specific contexts. In particular, region and economic structures may matter in order for access to have a positive effect, as might the quality of education and its relevance to job prospects (Brockhoff, Krieger, & Meierrieks, 2014). Individual-level mechanisms that lead toward conflict as well as the extent to which and how educational access plays a role as conflict continues are especially important to understand. More ethnographic research that examines the ways in which educational access matters to actors involved in conflict would be helpful in this regard. We next turn to the effects of conflict on educational access.
Conflict Reduces Access to Education
It is perhaps intuitive that conflict may reduce access to education. War destroys schools, damages school systems, and injures or kills students and teachers, with consequences for education. In addition, combatants may attack schools, rendering them unusable, and intimidate students and teachers to prevent them from attending (Burde, 2014; Glad, 2009; Human Rights Watch [HRW], 2006, 2010a, 2010b; O’Malley, 2010; Save the Children, 2013; Sommers, 2005). NGOs such as HRW and UNESCO have recently begun to document the most direct of these effects, gathering data on attacks on education (e.g., HRW, 2006, 2010a, 2010b, 2011; O’Malley, 2007, 2010). The Global Coalition to Protect Education from Attack was founded in 2010 to isolate attacks from other education issues and focus specifically on this phenomenon. 9
The type and intensity of the armed conflict affect educational access, as does baseline educational access before conflict. Drawing on available household-level data sets, several empirical studies investigated how exposure to conflict affects children’s schooling in order to understand long-term consequences for human capital formation (e.g., Akresh & de Walque, 2008; Dabalen & Paul, 2012; Shemyakina, 2011; Valente, 2011; Verwimp & Van Bavel, 2013). Specifically, these studies measured the direction and magnitude of the effects of violent conflict on educational attainment and its relationship to the employment status of those exposed to the conflict. The studies reached different, and sometimes contradictory, conclusions.
In general, conflict lowers access to education. A cross-country quantitative analysis of worldwide UNESCO data from 1980 to 1997 lent support to this conclusion, showing that war leads to a 1.6 to 3.2 percentage point decline in enrollment (Lai & Thyne, 2007). Whether this impact is greater for younger or older children, however, depends on context and time. Some studies have shown that conflict reduces younger children’s access to education more than older ones. For example, 4 years after the fighting ended in the Ivory Coast, children attending school in the most violent areas during the conflict had attained almost a full year less of education than those educated in the same areas before the conflict (Dabalen & Paul, 2012). Children in genocide-affected areas of Rwanda were significantly less likely to complete fourth grade and attained a year and a half less of school than those in other parts of the country. The impact was greater for children who were younger at the time of the genocide (Akresh & de Walque, 2008). Likewise, in Burundi, boys had a 6% lower chance of completing primary school for every year of exposure to conflict (Verwimp & Van Bavel, 2013). As in Rwanda, the impact was greater at younger ages among both male and female students.
Yet contrasting evidence shows that children who experienced conflict when younger may face greater barriers at higher levels of education. Longitudinal, mixed-methods survey, and interview data among children and youth in Sierra Leone showed that the likelihood of dropout increased with age because conflict reduced family income and children had to combine work and study (Zuilkowski & Betancourt, 2014). Similarly, the Bosnian conflict had a significant impact on secondary school completion, not primary education (Swee, 2009).
Interestingly, in some areas, the Maoist insurgency in Nepal increased female primary school completion, as well as male primary school enrollment (Valente, 2011). This outcome may be due to the Maoists’ emphasis on equality and the group’s efforts to remove barriers to schooling and to empower females. Nevertheless, female attainment decreased in areas of Nepal with high frequency of child abduction for fighting purposes (Valente, 2011). Likewise, demand for education may actually increase during conflict due to the desire to emigrate, as occurred in Spain’s Basque region (de Groot & Göksel, 2011). However, the nature and intensity of violence may be important to consider. Both the Basque country and much of Nepal are characterized by low-intensity, long-term violence. Finally, although educational attainment drops dramatically in fragile states, the decline appears to be more modest in conflict-affected states—a finding that is important to proponents of education as a tool for promoting state stability. Thus, state fragility (measured by the State Fragility Index) may be a stronger predictor of negative effects than conflict (Paulson & Shields, 2015; Shields & Paulson, 2015).
In part, variations in baseline enrollment rates may explain these contrasting findings on whether conflict reduces, increases, or has no effect on educational attainment. Paulson and Shields (2015) found that countries with high preconflict enrollment tend to see a decline in enrollment during conflict. However, countries with low preconflict enrollment rates continue to see slow increases in enrollment during conflict. 10 A variety of contextual factors as well as length and intensity of conflict mediate the impact of fighting on education. Although not examined in these studies, aid to education during conflict may also have a mediating effect.
Illuminating these findings, some qualitative studies have explored how conflict affects schools, making them unsafe for children as sites of forced recruitment (HRW, 2015) or kidnapping and extortion. Boko Haram’s mass abductions of school children in Nigeria are a high-profile example. According to interviewees, schools in Nepal have been targeted because they provided a physical space to house armed groups and also because they often had financial and human resources that armed groups can appropriate (van Wessel & van Hirtum, 2013). However, to date, there has been scant research on this topic, with qualitative researchers interested in the relationship between education and conflict more likely to pay attention to issues of curriculum and content, as elaborated below.
In sum, conflict is widely perceived to have negative effects on educational access, reducing the level of attainment for those who are exposed to violence. These data at the state level show clear deleterious effects. Yet cross-national comparisons reveal that state fragility may be a stronger overall predictor of educational decline in a country as a whole than conflict.
Conflict Reduces Girls’ and Boys’ Access to Education Differently
It is widely assumed that conflict affects girls’ education more than boys’, and that conflict likely exacerbates the same factors that inhibit girls’ access to education in peacetime. For example, breakdown in the rule of law and in social norms may increase parents’ concerns about a girl’s journey to and from school and about the threats to girls in school (Kirk, 2011). Although the obstacles that prevent girls from attending school may also exist in peacetime, available evidence has not compared the effect of specific obstacles in stable versus conflict-affected areas. Thus, we cannot quantify the different effects. However, qualitative assessment has shown that conflict deepens these existing concerns. Furthermore, substantial anecdotal evidence has indicated that once girls reach school, additional risks await. When social norms and the rule of law break down during conflict, girls may be at increased risk for sexual assault at school. One grade level may include children of different ages because their education may have been interrupted by violence. The presence of older boys in the classroom can be threatening to young girls and raise parents’ concerns (Kirk, 2009). However, these interactions—reports of which come primarily from practitioner experience—have not been empirically tested. Additional research should further test existing theory on why girls are less likely than boys to enroll in and attend school during armed conflict. Below, we discuss the scant evidence that does exist on girls’ education in conflict zones and how conflict affects access differently for boys and girls.
Existing empirical evidence has shown that girls’ enrollment is particularly sensitive to distance, which may be aggravated by social instability—more so than boys’. Girls’ attendance in Afghanistan decreased by 19 percentage points for every mile increase in distance from school (Burde & Linden, 2013). Social norms about propriety—for example, a belief that girls and women should stay out of public view—may affect girls’ mobility. Although many factors make it difficult for girls to travel to school in Afghanistan, political violence and safety concerns appear to exacerbate these barriers. Tajik women—but not men—who experienced conflict during 1992 to 1998 also attained less education than their peers in areas with less conflict (Shemyakina, 2011).
A handful of studies have examined why conflict may not have the same effect for boys. Conflict may increase the opportunity cost of time lost in travel to and from school for girls. For example, Somaliland mothers often worked outside the home to support the family because men had been killed or fled. Thus, they relied on girls to work at home more than they would otherwise (Bekalo, Brophy, & Welford, 2003). Some empirical research indicates that distance may affect girls’ access to education disproportionately even in peacetime, because a long walk provides opportunities to question girls’ sexual purity (Murphy, Stark, Wessells, Boothby, & Ager, 2011). Distance may combine with conflict to heighten fears that girls will be kidnapped or attacked, further reducing parents’ willingness to send girls to school (Burde & Linden, 2013).
Whether crisis affects male or female students more severely may depend on the setting, however. For example, male enrollment may be more affected by civil wars than female enrollment because of conscription (Lai & Thyne, 2007). This finding does not necessarily contradict the fact that girls face exceptional issues when it comes to schooling during conflict. One explanation is that girls typically have lower levels of enrollment to begin with (Kirk & Winthrop, 2007). Yet girls are not automatically worse off than boys on all measures. The two groups face different problems in different contexts that must be considered at the country or regional level. In the “What Works to Address These Issues” section below, we examine interventions designed to increase access to schooling in areas of armed conflict for girls in particular and children in general. Next, however, we look at the theories and empirical evidence on the relationship between conflict and learning.
Learning
Although education in emergencies has historically prioritized more easily quantifiable outcomes regarding access, with the shift in focus to fragile states scholars and international practitioners increasingly emphasize the need to evaluate the quality of learning in emergencies both to understand achievement as well as to understand how educational content and practices may contribute to conflict and instability. To date, much of the literature examining learning and achievement remains theoretical. In addition, the majority of scholarship that considers the relationship between education and learning focuses on content. For example, theorists point out that what is taught and how it is taught may increase the likelihood of conflict. Often these theories relate to materials on history or social studies, or to how a “hidden curriculum” socializes students into particular values. Few studies empirically assess the impact of educational content and teacher practices on conflict and vice versa. Existing literature also tends to conflate interpersonal conflict and societal conflict, implying, for instance, that abusive or oppressive classroom practices increase broader social tensions, but without spelling out how.
Educational Content Contributes to or Mitigates Conflict
What is the relationship between the content of education and likelihood of conflict? Although it is not clear to what extent education, either alone or in combination with other factors, contributes to conflict or peace, education-related factors almost certainly do contribute to participation or nonparticipation in violence. Both the explicit curriculum, such as what textbooks say about a particular group, as well as the hidden curriculum, which may teach values and norms implicitly, are important for their impact on societal conflict (Barakat & Urdal, 2009; Collier & Chauvet, 2007; Collier & Hoeffler, 2004).
Indeed, history and historical detail can be easily manipulated in textbooks, increasing social tensions by emphasizing the historical narrative of one group over another or actively portraying another group in negative and threatening terms (King, 2011, 2014; Kirk & Winthrop, 2007; Paulson & Rappleye, 2007; Smith, 2005). This work focuses on societal conflict, with the implication that promoting biases and stereotypes may also shape interpersonal interactions. For example, if a child learns to identify with an ethnic or religious group in school, and is taught that another group threatens his or her group, the child may be more likely to participate in or support violence against the other group. These lessons may be particularly problematic because teachers and textbooks are highly respected in most societies, particularly in those that are print-poor, and children may be unlikely to question what they learn (Burde, 2014). Conversely, textbooks that include respectful depictions of different groups, avoid stereotypes, and include diverse ethnic groups rather than disproportionately representing one group are likely to correct social biases (Dryden-Peterson, 2010; Dupuy, 2008). Such changes may reduce the potential for conflict: Students in the dominant group may be less likely to believe negative portrayals of the minority if textbooks and teachers model respect and tolerance, and the minority group may feel less alienated and have less motivation for conflict if they are respected in school (Wedge, 2008).
Recent theoretical work has begun to consider why some peace education interventions that emphasize intergroup contact may have mixed results—although these hypotheses have not been tested. Zembylas and Berkman (2015) noted that standard integration programs often reify identity differences, assuming that groups are relatively homogenous and stable. Furthermore, they tend to minimize poverty and power, ignoring how these factors may influence students’ understandings of experiences and histories. Critically, integration programs often create a “bubble” of contact, and so are unlikely to address structural inequalities to which the students return. Furthermore, peace education is primarily a Western construct and so may not work in poorer conflict-affected environments (Sommers, 2001). Finally, although the theories described above and in most examples given below focus on social (group) conflict, peace education interventions are implemented and analyzed at the individual level. We discuss effects of peace education interventions in more detail in the “What Works to Address These Issues” section below.
In addition, theoretical articles consider whether values and assumptions that are often embedded in school lessons (the hidden curriculum), may influence interpersonal conflict by shaping individuals’ actions and behaviors. During wartime, classes may be more susceptible to violence and corruption due to the use of physical punishment, curricula that teach about war, or cheating in examinations (Davies, 2005, 2008). Violent or aggressive teaching techniques that model bad behavior likely make a child more violent outside the classroom. Alternatively, teachers may model conflict resolution skills by prohibiting violence in class and creating an atmosphere of open and critical discussion (Davies, 2004, 2008, 2014). Similarly, teaching a tolerant, inclusive conception of citizenship, and treating all students equally may decrease the likelihood of discrimination. We now consider the empirical links between negative classroom content and increased risk of conflict. Studies of the positive effects of learning on conflict mitigation focus on interventions, which we discuss in the next section.
Almost all of the empirical studies that were returned in our search and that relate to this section examined divisive educational content, implying but not showing that what is taught fuels conflict or how it might do so. It also examined mainly the relationship between societal—rather than interpersonal—conflict and education. Indeed, these studies showed that many states have used curricula and textbooks to promote social divisions, both at home and abroad. Collecting extensive interview data and analyzing textbooks that were in circulation prior to the 1994 genocide in Rwanda, King (2014) showed how the history of Hutus and Tutsis was portrayed as inherently oppositional. Textbooks taught that Hutus had been slaves to Tutsis and that these relationships had continued into the present. By teaching children to identify against another group, these texts helped create an atmosphere conducive to violence in Rwanda.
Additionally, a hidden curriculum likely increased societal violence. Interviewees identified practices that they believe helped fuel later conflict: Teachers were aggressive, made negative comments toward one ethnic group, and tolerated harassment by students of that ethnic group. Interviews also suggested that valuing obedience over critical thinking in school may have indirectly facilitated participation in conflict (King, 2014). Similarly, Pakistani textbooks have been shown to underscore sectarian divisions (International Crisis Group, 2014), and India and Pakistan intentionally devised curricula that inculcated exclusivist identities denigrating the other (Kumar, 2002; Lall, 2008). A systematic analysis of both countries’ textbooks showed that elements of history were omitted or emphasized to cultivate an exclusivist Hindu or Islamic identity, respectively. 11 Likewise, during the civil war in Sudan, textbooks and teaching materials, particularly in the north, emphasized exclusivist identities, likely contributing to conflict (Breidlid, 2010). The perception of biased schooling may also fuel violence. A sense that public schools are agents for Christianization has been found to increase alienation of Muslims on the Philippine island of Mindanao, contributing to wider social conflict (Milligan, 2003).
Evidence has shown that language of instruction is also contentious in relation to societal conflict. Similar to the textbook and curriculum studies, this research assumes that divisiveness directly incites conflicts or fuels ongoing violence. For example, groups may be disadvantaged if they do not speak the language in which classes are taught, deepening social inequality and potentially intensifying grievances. Extensive fieldwork in six Burmese refugee camps on the Thai-Burmese border showed that schooling was conducted in a language that one quarter of the population in the camps—primarily Muslims—did not speak and that textbooks were written in English (Oh & van der Stouwe, 2008). The schools privileged the dominant group: “Western-oriented, mostly Christian, and Skaw Karen-speaking” (Oh & van der Stouwe, 2008, p. 611). Similarly, the official use of English in Rwandan schools may create new inequalities because it benefits the Tutsi, who more likely have English-language backgrounds (King, 2011).
However, it is also difficult to establish a causal link between language policy and violence. Furthermore, these studies have not isolated language policy among other potentially relevant factors (e.g., economic position) motivating conflict. They have suggested, rather, that language policy may play a role in motivating resistance to a dominant group, and language policy may indeed contribute to conflict even if it is not the only motivating factor. Furthermore, the findings described in the preceding paragraphs assume that what and how students learn in school informs how they behave outside school. The process is a complicated one that does not always produce clear results.
In contrast to the studies above that analyzed group behavior, studies that analyzed individual behavior note that not all students who are exposed to violent or exclusionary messages behave violently. A series of interviews with students living in Eastern Europe and Germany during World War II and in Afghanistan during the post-1979 armed conflict, for example, showed that students may resist such messages. Although interviewed after the fact, interviewees often expressed remorse over the divided identities they learned in school, “suggesting internal resistances” to these messages (Dicum, 2008). Data based on such historical memory can be fallible and subject to selective reporting (King, 2014). Yet similarly, although jihadist textbooks widely used in Afghanistan and in Afghan refugee camps in Pakistan likely had an effect on students’ attitudes, many students exposed did not become jihadists (Burde, 2014). Thus, researchers and policymakers should not assume that children inevitably believe everything they are taught, that what students learn is not subject to later reevaluation, or, at a minimum, that attitudes translate into violent action.
In sum, various forms of inequality and discrimination manifested in the classroom and classroom materials are thought to drive violence, and critical thinking skills are theorized to help resist violent messages, but more research is crucial to identify and understand these mechanisms, distinguish between group- and individual-level mechanisms, and understand the links between the two. Generally, more research is needed on the potential impact of curriculum, language policy, and teacher practice on violence, though there is a large body of research that examines how curricula in the form of peace education may or may not mitigate conflict. We address this research in the “What Works to Address These Issues” section below.
Impact of Conflict on Educational Content and Learning Outcomes
There is likely a reciprocal relationship between curriculum and conflict—negative and biased curriculum fuels conflict and conflict increases the chance that curriculum will be biased. Regarding the latter, Davies (2004) has suggested that schools are unlikely to offer peaceful environments characterized by a respect for children’s rights if the wider society is repressive or conflict-affected. We discuss the very limited evidence on this latter relationship here.
A handful of studies have suggested that the political context after conflict affects how an open and safe environment in the classroom as well as textbook material develop (Freedman, Weinstein, Murphy, & Longman, 2008). For example, the governments of Rwanda and Bosnia both prohibited teaching history directly following their conflicts. Such a moratorium is controversial and there is no clear evidence that suggests the benefit outweighs the possible cost. In Rwanda, the official government narrative, which has emphasized unity and discouraged the examination of past or current conflicts, may be problematic. Teachers have remained afraid to discuss ethnicity in the classroom, and recent oppressive government policies seem to have further inculcated a climate of fear. The suppression of classroom debate and analysis of conflict likely have stifled the development of critical thinking skills and the understanding of historical evidence. Furthermore, ignoring students’ and teachers’ experiences of ethnic difference and conflict could lead to more conflict in the future (King, 2014). Over the course of a history curriculum reform project, the Rwandan government became increasingly less willing to support the development of innovative curricula and teachers backed away accordingly (Freedman et al., 2008). Thus, conflict may negatively affect curriculum content even after the violence ends.
Practitioners recognize that, in situations of conflict, schools lack adequate pedagogical equipment and that the learning conditions are often abysmal (Pingel, 2010). They intuitively link these conditions to reduced levels of learning. However, we found no scholarly research examining this relationship. Additional research is required to understand the extent to which conflict negatively affects children’s learning in relation to other factors, such as poverty. Existing research pays more attention to other risks, such as to students’ psychological well-being. Next, we assess efforts to promote protection and well-being through education.
Protection and Well-Being
The psychosocial impact of violence and exploitation on children and teachers is less visible, but may be no less damaging than physical injuries. Most scholarship considering the relationships among education, conflict, and psychosocial well-being is theoretical in considering, on one hand, the impact of trauma on children’s education and, on the other hand, the potential for schooling to mitigate the effects of this trauma. Recent research, however, indicates that daily stressors play a more significant role than trauma (Betancourt et al., 2013; Burde et al., 2015; Fernando, Miller, & Berger, 2010; Wessells & Kostelny, 2012).
The Protective Potential of Education
International practitioners consider it generally evident that traumatic stress symptoms among children in these settings interfere with their learning (e.g., Aguilar & Retamal, 2009; Muñoz, 2010). Empirically, however, the effects that conflict trauma has on children’s learning have received limited attention beyond a few isolated studies. Just two studies in Sri Lanka and Sierra Leone have shown that learning is impaired by posttraumatic stress (Betancourt et al., 2008; Elbert et al., 2009). The lion’s share of literature focuses on how conflict trauma affects children writ large.
Conversely, educators firmly believe that education protects children from harm. Indeed, this is a key assumption held by those in the education in emergencies field and used to justify the humanitarian provision of education. International organizations have invested significant resources to develop child-friendly schools/spaces, and INEE Minimum Standards highlight this claim: “In emergency situations through to recovery, quality education provides physical, psychosocial and cognitive protection that can sustain and save lives” (INEE, 2010, p. 2).
This perspective assumes that education mitigates the effect of conflict on children in several ways. First, perhaps most prominently, schooling is thought to provide children and teachers with a sense of normalcy in an otherwise disruptive setting by giving them a structured and stable routine (Davies & Talbot, 2008). Next, trauma can be addressed in the school setting by giving children targeted psychosocial support and developing their natural resiliencies. For example, the provision of educational activities in a “healing classroom” may contribute to the positive social and emotional adjustment of children (Winthrop & Kirk, 2008). 12 Finally, schools are considered entry points for providing children and communities with access to other life saving services, such as nutrition, health, or water and sanitation (Aguilar & Retamal, 2009; Mosselson et al., 2009; Paulson & Rappleye, 2007; Smith, 2005).
There is mixed empirical research to bolster these claims. A longitudinal qualitative study of refugee children showed that education may promote both immediate physical security and enable children to hold aspirations for future economic security (Dryden-Peterson, 2011). Other studies indicate how children and students themselves value schooling to support their well-being. Interviews with refugee youth illustrated the importance that students themselves place on different types of learning, from technical knowledge to practical skills to critical thinking (Winthrop, 2011). A qualitative study of an IRC nonformal education and recreation program for Chechen IDP (internally displaced person) children involved young people as youth leaders who cooperated with teaching staff to engage with and develop activities for other young beneficiaries. Although the program provided a space to develop important social relationships and engage in meaningful activity, the fact that the education program was nonformal limited the youths’ ability to develop the sense of normalcy that they wanted (Betancourt, 2005). In the section on “What Works to Address These Issues,” we examine education-based psychosocial interventions specifically designed to promote children’s well-being and protection in armed conflict.
Constraints on the Protective Potential of Education
International practitioners also point out several constraints on the protective potential of education. In particular, as mentioned above, schools may be a prime target for combatants to attack (HRW, 2006, 2010a, 2010b; O’Malley, 2007, 2010)—something that negates the safe space that schools may provide (Aguilar & Retamal, 2009). Additionally, teachers themselves are often traumatized, with negative repercussions for the classroom. Girls may also be at increased risk of sexual assault, harassment, and exploitation at school during conflict (Kirk & Winthrop, 2006; Nicolai & Triplehorn, 2003). Finally, simply returning to normal classroom activities may be insufficient among children who face difficulties concentrating after exposure to violence and disruption (Aguilar & Retamal, 2009).
Despite a significant gray literature that proposes ways in which education promotes the protection of children in situations of armed conflict, few studies specifically examine how this type of psychosocial protection might improve educational outcomes. Neither does the literature examine other educational dimensions like classroom climate and interpersonal relationships that might indirectly affect both academic outcomes and psychosocial well-being—although one study doing so is in process (see Torrente et al., 2015).
What Works to Address These Issues
In this section, we examine evidence on what types of interventions might work to mitigate the potential negative relationships between education and conflict and improve educational outcomes in conflict situations. We include in our discussion evidence from intervention studies that employ RCTs, also referred to as impact evaluations, as well as from other kinds of empirical research such as rigorous observational studies and quasi-experimental designs. We first discuss programs that increase access to education in situations of conflict. Next, we review educational interventions intended to mitigate conflict through their content. Finally, we examine protection-oriented interventions designed to bolster education and well-being. It is important to note that there may be many other interventions that are effective but are not included here because of limited available evidence assessing their impact.
Access
In contrast to a relatively substantial literature on the relationship between access to education and conflict, little research has tested the impact of interventions intended to increase access in conflict-affected settings. In other words, there are few studies that have explained what works to remedy the access issues described above. Most studies of interventions designed to enhance access have been descriptive or process-oriented case studies conducted by the international agencies providing education and have neither been systematic nor explanatory. Here, we review several studies that meet our selection criteria.
Community-based schools
Existing evidence has shown that community-based schools increase access to education, particularly by decreasing distance to schools and encouraging participation of local communities. According to one impact evaluation, community-based schools, supported by international NGOs, increase enrollment particularly for girls in conflict-affected Afghanistan. Although attendance increased for both boys and girls dramatically, rising to 70% from an average of 27%, girls benefitted the most from the schools. Their enrollment increased by 15 percentage points more than boys’, and girls’ test scores increased by 0.25 standard deviations more than boys’. Furthermore, the gender gap in test scores narrowed significantly in the treatment villages (Burde & Linden, 2013). These schools were extremely effective at increasing educational access and reducing gender inequality.
Community-based schools may not only improve access to education but also provide protection by deterring attacks that target both girls’ and boys’ education (Burde, 2014; Rowell, 2013). Data from the NGO CARE on the rates at which armed groups attacked government and community-based schools indicate that, because they are located in a home or mosque and are more easily monitored by the community, community-based schools may offer less of a target in areas where government schools are under attack (Burde, 2014; Rowell, 2013).
Girls’ schools and female teachers
The availability of girls’ schools, female teachers, and/or female classroom attendants also increases girls’ access to education. In Pakistan, the presence of a girls’ school in a village significantly increased the likelihood that girls from that village enroll (Lloyd, Mete, & Sathar, 2005). In Afghanistan, according to observational data, as the proportion of female teachers increased, girls’ enrollment rose by 30% (Guimbert, Miwa, & Nguyen, 2008). This may break a negative cycle: Too few female teachers keep girls from attending school, and too few educated girls cannot produce enough female teachers.
However, many communities stress the importance of personally knowing their daughters’ teachers. Both experimental and anecdotal evidence has indicated that if parents know the male teachers who instruct their daughters, the teacher’s gender is not a deterrent to girls’ enrollment, at least before puberty (e.g., Burde & Linden, 2013). Additional research should disaggregate gender from acquaintance to test the effects of both on girls’ enrollment.
In sum, supporting community-based education is a common intervention in countries affected by crises, with clear and significant effects on children’s access to school and learning. Beyond these initiatives, surprisingly, there is little empirical work that specifically examines interventions to increase access to education in crisis settings. Most international evaluation research examines aid to education in peaceful low- or middle-income countries. Also surprisingly, we found no impact evaluations to date of temporary learning spaces or accelerated learning programs, despite how common these interventions are in crisis situations. Finally, more research studying how education interventions can protect children is necessary.
Learning
Programs that aim to address educational content and practices are often designed to contribute to peacebuilding after conflict (Davies, 2005; Kagawa, 2005; Paulson & Rappleye, 2007; Smith, 2005). Much of the intervention research examining the peaceful potential of education has focused on these programs. Among these, peace education interventions and multiple-perspective history teaching have been evaluated rigorously. As mentioned, these evaluations analyze individual-level outcomes, assuming that reducing personal biases and interpersonal conflict will mitigate social conflict. An overarching limitation of these interventions and this literature, therefore, is that they do not specify how interpersonal and societal conflict are linked nor analyze the impact of peace education on group outcomes.
Peace education programs
Peace education is generally premised on the idea that education can promote conflict resolution and peace-building by fostering an atmosphere of nonviolence and reconciliation among learners that will diffuse into the community, often—though not exclusively—through direct contact between groups (Allport, 1954). Additional components of peace education include activities and lessons on sharing and dealing with emotional stress positively (Webster, 2013). Because substantial literature exists on peace education, we provide only an overview of key findings here, building on the earlier work of Bar-Tal and Rosen (2009; also see Burde et al., 2015).
Intergroup contact between conflicting groups is often an important component of peace education programs. Although there is limited evidence on the long-term effects of intergroup contact, one impact evaluation of a peace workshop in Sri Lanka found that effects lasted up to a year after the intervention ended (Malhotra & Liyanage, 2005). The workshop consisted of a 4-day camp where 18- to 21-year-old Sinhalese and Tamil students lived and socialized, attended minilectures, creative activities, and toured a multiethnic village. In behavioral tests 1 year later, participants expressed greater empathy toward members of the other ethnicity and were more willing to donate money to charities that benefited the poor of the other ethnicity. Meanwhile, students in the control group, as well as students from nonparticipating schools, showed less empathy and were less likely to give to needy members of the other group.
Combining independent study with intergroup contact also shows some promise. For example, a year-long school-based program that fostered intergroup contact between Palestinian and Israeli youth found that both groups placed more emphasis on positive aspects of peace such as cooperation and harmony (Biton & Salomon, 2006). The program also prevented the deterioration of views and feelings among program participants as compared with the control group in the face of adverse events. The fact that the program was long-term, executed by trained classroom teachers, and aimed at breaking down stereotypes may have contributed to its success. These mechanisms need to be tested, however. In addition, it is difficult to know whether the effects were lasting, as the last survey was conducted immediately after the program ended.
Peace education often positively affects attitudes and perceptions in postconflict, protracted or intractable conflicts among willing participants (Bar-Natan, 2004, as cited in Salomon, 2004; Burde et al., 2015; Maoz, 2000), yet it is unclear whether these positive outcomes are retained over time. A quasi-experimental study of a year-long, school-based program for Israeli Jewish and Israeli Palestinian youth found changes in adolescents’ peripheral beliefs, but not the core beliefs that were central to their groups’ collective narratives (Rosen & Salomon, 2011). A second posttest 2 months later showed that even the peripheral beliefs changed back to their initial states.
In sum, although the interventions described above show promise, most evaluations of peace education programs have struggled to isolate positive changes to the curriculum or classroom practice from among the other factors that could help build peace. In addition, the outcome measures typically used in evaluations have been insufficient for assessing these programs, which often seek to change the “root causes” of conflict, by changing attitudes and values. Lack of comparability between experimental and control groups and selection bias among participants have also weakened most of these studies. Additional research including experimental designs should seek to better understand the mechanisms that promote the outcomes that peace education programs are designed to achieve.
Multiple-perspective history teaching
Evidence from intervention studies has shown that, compared with conventional history teaching that focuses on a single narrative, teaching multiple perspectives of history positively affects students’ attitudes and perceptions (Burde et al., 2015). Practitioners promote these perspectives in two key ways: (a) an inquiry approach to multiple-perspective history teaching (also called “critical analysis of conflicting sources”) and (b) a dual narrative approach to history education (also called “joint history textbooks”).
The first approach assumes that engaging critically with primary and secondary evidence on conflicting historical perspectives will transform conflict-ridden societies (McCully, 2012). Existing experimental and quasi-experimental evidence is promising. Comparison of the inquiry approach with conventional and other multiperspective approaches with history teaching showed that the former had a significantly positive impact on changing students’ attitudes and perceptions about the contested past and out-group members (Barton & McCully, 2010; Goldberg, 2014) and on improving intergroup relations (Goldberg & Ron, 2014).
The dual narrative approach, in which students learn both in-group and out-group historical narratives, has mixed results. For example, some interventions in which conflicting groups jointly produced dual narrative histories were promising in the Israeli–Palestinian context (Adwan & Bar-On, 2004; Maoz, 2000). However, dual narrative textbooks used by Palestinian high school students in Israel did not promote interest in the Israeli perspective and were associated with negative reactions to the latter’s narrative (Eid, 2010). In Northern Ireland, students subjected to a dual history curriculum showed more interest in out-group narratives, but struggled to fully engage with alternative historical perspectives (Barton & McCully, 2012).
Most of these programs address intergroup violence that typifies a civil or local war, such as the conflicts in Northern Ireland or between Israelis and Palestinians. They generally work with young people who agree to participate; they do not address internationalized conflicts or attempt to prevent recruitment into violent extremism. Much work remains to be done to understand whether, how, and to what extent education programs can mitigate the phenomena that trigger participation in violent extremist groups and activities.
Protection and Well-Being
In the nexus among education, conflict, and protection, most research on education and protection has assessed formal and informal education interventions designed to mitigate anxiety, depression, or trauma in these circumstances. Little research has examined how psychosocial programs may improve educational outcomes. Although interventions adopt a range of approaches—from addressing trauma to promoting resilience—scholars debate the appropriateness of using clinically oriented approaches among a general population. In particular, the posttraumatic stress disorder (PTSD) pathology established in the West is likely incompatible with many local beliefs elsewhere (Boothby, 2008; Burde et al., 2015; Fernando et al., 2010). Interventions that do not take local norms and customs into account may be inadvertently harmful, and too much focus on trauma may overlook the important role of postconflict stressors on mental health (Betancourt et al., 2013; Fernando et al., 2010; Wessells & Kostelny, 2012). Here, we review one experimental study assessing how a psychosocial intervention improves children’s educational outcomes, along with other school-based and non–school-based interventions designed to address trauma and promote well-being. These programs typically provide creative arts or drama therapy, psychosocial support, and recreational activities.
These studies often have investigated whether an intervention is associated with a reduction in PTSD and, to a lesser extent, with depression and anxiety. Although the intervention studies reported here meet our criteria for selection, as with much of the research in this emerging field, there are some caveats to keep in mind when reading these outcomes. First, because these studies typically have not isolated program components, we do not know specifics regarding what works and what does not in achieving targeted outcomes. Second, long-term follow-up is unusual and often challenging making it impossible to assess the effects over time. Third, most studies have depended on self-reports, not supported by behavioral tests, observations, or teacher/parent reports of child functioning. Finally, most have conducted pre- and posttreatment tests, not accounting for selection bias, nor capturing changes over the course of the intervention. That said, overall these studies have found promising results in reducing children’s anxiety, depression, and trauma. And taken together, they offer important insights and points of departure for future research.
The following is a description of clear successes or controversies that have emerged in the course of psychosocial interventions. In this discussion, we treat psychosocial support programs as relatively monolithic, assessing the evidence on their outcomes. This is because of the difficulty in understanding what program components contributed to which outcomes and lack of attention to causal effects.
Psychosocial therapy may improve educational outcomes
Our search returned only one study that examines the impact of a psychosocial intervention on children’s education; however, its results are promising. Using an RCT to examine the impact of a cognitive-behavioral therapy (CBT)/group therapy intervention and education subsidies, researchers randomly assigned war-affected youth (aged 15–24 years) to receive CBT through 10 group sessions (Betancourt et al., 2014). During these sessions, the youth discussed self-concept and the impact of trauma on interpersonal relationships. Immediately following treatment, the youth were again randomized and offered an education subsidy. Enrollment, attendance, and classroom performance were assessed among in-school youth after 8 months. The CBT intervention had significant positive effects on both psychosocial indicators and educational outcomes, including enrollment, attendance, and classroom behavior. Significantly, the education subsidy alone increased attendance, but did not have an effect on mental health, school retention, or classroom behavior. Furthermore, interactions between the therapy and the subsidy were not significant, thus the psychosocial support was the primary mechanism improving educational outcomes.
School-based psychosocial interventions
The range of psychosocial interventions that take place at school is diverse, but they include classroom-based or group interventions inside the school but outside the classroom. Although the differences in the effectiveness of school-based programs versus clinical interventions are not clear, school-based programs allow practitioners access to a larger number of children and make treatment more accessible. Interventions that combine education programs with trauma healing techniques or that integrate healing activities in a school curriculum are rare; we review the available research that meets our criteria.
Some existing evidence shows some positive results on whether providing psychosocial support to children in school through trauma-focused activities helps them recover from crisis. Trauma/grief-focused psychotherapy provided to Bosnian secondary school students exposed to war significantly reduced depression, grief symptoms, and PTSD according to self-reports (Layne et al., 2008). The intervention consisted of cognitive-behavioral activities that included cognitive restructuring, problem-solving skills, psychosocial education, relaxation skills, stress management, and therapeutic exposure. Reductions in PTSD and depression symptoms were connected to increased classroom compliance and interest in school and negatively associated with school anxiety and withdrawal. Likewise, a curriculum-based training program with fourth and fifth graders in Croatia included trauma healing elements, as well as awareness of ethnic bias, education on human rights, and instruction in nonviolent conflict resolution (Woodside, Barbara, & Benner, 1999). The results showed a reduction in PTSD symptoms in the intervention group, as well as a small but significant reduction in ethnic bias.
Similarly, school-based psychosocial interventions that integrate art and recreational activities show improvement in mental health. A classroom-based program, “Overshadowing the Threat of Terrorism,” was implemented with elementary Israeli students who were randomly assigned to either the intervention or the control group (Berger, Pat-Horenczyk, & Gelkopf, 2007). The intervention combined cognitive-behavioral activities with art and recreational activities, as well as with self-expressive activities in the curriculum. Two weeks after the program ended, participants reported statistically significant reductions in PTSD symptoms and functional impairment measures compared with those in the control group; however, the long-term impact of the intervention was not assessed. The Psychosocial Structured Activities used with war-affected youth in Northern Uganda appeared to have similar effects (Ager et al., 2011). An intervention consisting of didactic presentations, drama, games, and reflective exercises, as well as community service and periodic facilitated discussions between student participants and their parents was implemented in 21 conflict-affected schools. For a 12-month follow-up evaluation, eight schools were selected by random quota sampling from those receiving the intervention and students were assigned to intervention and comparison groups. Although both students who participated in the program and those who did not showed a significant increase in their well-being, the well-being of children in the group receiving the intervention increased significantly more than for those who did not. A systematic review of 21 studies on interventions targeting refugee children (14 in high-income countries, 7 in refugee camps) also indicated the effectiveness of creative arts–based programs in significantly improving mental health (Tyrer & Fazel, 2014). However, the review also showed that interventions using CBT had the largest effect sizes.
Yet significantly, school-based psychosocial interventions may not be equally effective for all populations. A rigorous randomized evaluation conducted with Palestinian schoolchildren (aged 10–13 years) showed weak effects for girls. Exploring the effects of a psychosocial intervention using “Teaching Recovery Techniques” on emotional regulation, researchers found that whereas PTSD symptoms were reduced for boys, symptoms were reduced in only a subset of girls: those who had lower levels of trauma at baseline (Punamäki, Peltonen, Diab, & Qouta, 2014).
Training teachers to cultivate protective learning spaces
Some evidence from interventions that train teachers to cultivate protective environments in their classrooms shows mixed results. A study of an intervention in Israel conducted after the 2006 Lebanon war found that when teachers established a safe environment using diary documentation, narrative techniques, and play activities to help children reprocess traumatic experiences, children’s coping skills improved (Wolmer, Hamiel, Barchas, Slone, & Laor, 2011). Overall, participating children were more likely to maintain a healthy equilibrium and re-experience traumatic events less frequently. Yet because data were collected in pre–post and follow-up assessments of the children, there is no counterfactual to which to compare these outcomes. A partial evaluation of IRC’s Healing Classrooms intervention in the Democratic Republic of Congo assessed the effects of training teachers to create safe and supportive learning environments (Torrente et al., 2015). After the first and second waves of data collection, the researchers found improved student perceptions of support from their schools and teachers among language-majority students, but not among language-minority students. However, the program also reduced students’ perceptions of their schools as predictable and cooperative. There were no effects on student self-perceptions of well-being.
Informal psychosocial interventions
Setting up child friendly spaces/child-centered spaces or temporary learning spaces are common interventions implemented by humanitarian agencies working on child protection and education in the weeks following the acute onset of conflict. The two often look similar: Both often take place in tents and combine lessons and recreational activities with the goal of providing structure and a safe space for children in the midst of chaos, although temporary learning spaces are typically more focused on education.
Existing evaluations of this type of program have shown positive outcomes. One evaluation found that children who participated in a child-centered space in an IDP camp in postconflict Northern Uganda had better psychosocial outcomes than those in a nearby camp that did not receive the intervention (Kostelny & Wessells, 2008). Participants were more likely to be safe at home and in the community, had positive social interactions with their peers and adults, and learned stronger life skills than those who did not participate in the child-centered space.
However, factors external to schooling—specifically caretaker mental health and attitude—appear to be important as well. An emergency education intervention among refugee youth living in Kunama camp in Ethiopia involved educational, psychosocial, and recreational activities conducted in shifts in large tent schools, as well as recreational clubs when school was out of session (Betancourt, Yudron, Wheaton, & Smith-Fawzi, 2012). Drawing on data from surveys conducted 1 year apart with both adolescents and caregivers, the authors found that caregiver mental health and attitudes were significant in determining how positive an effect an emergency education program had on youth mental health and behavior. When caregivers had higher levels of distress, youth were more likely to have higher levels of psychosocial trauma and behavioral problems. Conversely, when caregivers had lower levels of distress, perceptions of good access to social services, and higher socioeconomic status, youth had fewer behavioral problems. Similarly, an experimental study from Bosnia tested a 5-month, nonformal education intervention that consisted of weekly meetings to promote good mother–child interaction, increased knowledge of child development and trauma, peer support, and basic health care (Dybdahl, 2001). Psychological tests and qualitative observations revealed significant postintervention differences between treatment and control groups on measures of maternal well-being and mental health, as well as on child psychosocial functioning.
In sum, the protective potential of education is a key message from education in emergencies practitioners. There is some evidence that psychosocial support at school helps children recover from crisis through improved educational outcomes and improved mental health—although these effects may be reduced by factors external to the psychosocial interventions. Furthermore, the extent to which formal education itself provides protection outcomes has not yet been rigorously tested. This is an important area for future research.
Discussion
The increasing demand for more specific evidence on program effectiveness over the past several years is already beginning to show the new directions in which the field will move. Peace education is now its own subfield, as is the study of psychosocial interventions. Alternative learning systems such as accelerated learning programs, community-based education, and temporary learning spaces are developing new schools of thought. We summarize below a few conclusions regarding what we know about the relationship between education and conflict, and how to support education in emergencies in light of these trends.
Access to Education
It is clear that conflict affects access to education, with school infrastructure destroyed during conflict and school personnel killed (Cuaresma, 2010; Lai & Thyne, 2007). The impact tends to be greater for younger children than for older ones (Akresh & de Walque, 2008; Dabalen & Paul, 2012; Verwimp & Van Bavel, 2013). However, in certain contexts the negative impact of fighting on education is not as clear (de Groot & Göksel, 2011; Swee, 2009; Valente, 2011) and may be mediated by a variety of contextual factors as well as the intensity and length of conflict (Shields & Paulson, 2015). More evidence is needed on these mediating factors.
We can also say with some confidence that unequal access to education contributes to conflict initiation (Østby & Urdal, 2010; Thyne, 2006). In particular, there is rigorous evidence that youth without access to education are more likely to join violent conflict than those who attend school (Humphreys & Weinstein, 2008). This relationship appears strongest in low- and middle-income countries (Barakat & Urdal, 2009). More research is important to understand the mediating role of factors such as income level, economic structures, and geographical region that play a key role in this relationship (Barakat & Urdal, 2009; Collier & Hoeffler, 2004; Collier & Chauvet, 2007; Østby & Urdal, 2010). However, the possible link between education level and participation in or support for violent extremism has not yet been clearly established.
Furthermore, convincing evidence shows that distance to school is an important factor in determining enrollment in primary school, especially for girls (Burde & Linden, 2013; Guimbert et al., 2008). In addition, an increased proportion of female teachers is likely to encourage girls’ enrollment (Guimbert et al., 2008; Lloyd et al., 2005). Open questions include, for example, the conditions in which equitable access to education enhances peace. In addition, studies should illuminate at which levels of education—primary, secondary, tertiary—interventions are most effective in promoting peace. It is important to learn how the effectiveness of specific interventions varies according to level of country development (Brockhoff et al., 2014), or political dissatisfaction (Shafiq & Sinno, 2010).
Learning
A number of interventions are effective in rebuilding and maintaining education quality during and after crises. First, community-based education shows clear and significant effects on raising children’s achievement levels when available formal schools cannot reach war-affected school-age children, particularly girls (Burde & Linden, 2013). Second, contract teachers, although often less qualified than those with full government credentials, are critical to providing education during crises when many qualified teachers have fled or been killed.
Existing research does not establish a strong causal relationship between education curricula content, language policy, teacher practices, and violence. However, in examining the effect of particular curricula on attitudes and behaviors related to peace or conflict, it seems clear that classroom materials that frame particular social groups in a negative light, as well as repressive pedagogy, are likely to have negative outcomes on peace-building measures (Burde, 2014; Davies, 2004; King, 2014). In order to understand how this happens and to what extent this may be true, these mechanisms need to be tested systematically. Similarly, exclusions based on language of instruction likely have negative effects both on educational achievement as well as on positive indicators of peace, but these questions should also be examined with more nuance. In both cases, it would be valuable to understand more precisely and systematically the degree to which positive curricula and curricula that promote critical thinking produce positive effects.
Peace education programs show promising results in changing attitudes and behaviors toward members of those perceived as the “other,” at least in the short term (Biton & Salomon, 2006; Malhotra & Liyanage, 2005). However, more research is required to understand the long-term effects. In addition, it is not clear which program components lead to which outcomes.
Protection and Well-Being
With regard to the potential of education to offer protection and enhance well-being for children living in emergency and postemergency situations, research shows that providing children with structured, meaningful, and creative activities, in a school setting or in informal learning spaces, improves their emotional and behavioral well-being (Ager et al., 2011; Berger & Gelkopf, 2009; Kostelny & Wessells, 2008). In addition, some psychosocial support programs appear to improve classroom-related outcomes (Betancourt et al., 2014; Layne et al., 2008; Woodside et al., 1999) although further research is needed. New work should examine mechanisms that produce greater resilience among children as well the effectiveness among different psychosocial programs.
Conclusion: The Challenge Ahead
The review above shows the current state of research on education in emergencies. Yet this field is changing rapidly and world events have expanded attention to education in these contexts in a way that has never before occurred. These changes will continue as new voices introduce their ideas for peace and as events raise questions about the links between education and violent extremism. For example, the award of the Nobel Peace Prize to Malala Yousafzai increased attention to education in emergencies. Although still young, Malala already demonstrated that she is a force on the world stage—one that can amplify the voices of tens of thousands of civil society actors who argue for education as a human right and view it as an antidote to conflict. At the same time, strong states are examining the relationship between education and conflict with an escalating sense of urgency. In response, attention to education in emergencies is set to grow exponentially. Indeed, world leaders including the Prime Ministers or Presidents of Chile, Indonesia, Malawi, Norway, and UNESCO’s Secretary General have co-convened an Education Commission led by a former British Prime Minister that is tasked with defining and addressing the educational needs of children in low-income and conflict-affected countries and regions worldwide. These responses hold promises and pitfalls. The latter are more likely if interventions are not grounded in research and data collection.
Yet robust evidence on which to act is limited. Conducting rigorous research in conflict-affected environments is difficult and often complicated by the fact that most of the lead researchers of these studies to date are foreign to the contexts in which they are working. In addition, researchers launching projects in this field must contend with shifting geopolitical stakes and interests. Increasing commitment to rigorous research, circulating studies as they emerge, and coordinating the efforts of large actors with small, local actors and membership organizations will be critical in the years ahead.
The conceptual approach that dominates the field will also play a significant role in shaping the work ahead. The education in emergencies framework that has influenced the field to date highlights issues important to the rights of beneficiaries of these programs. In contrast, a framework that focuses on state fragility, which is increasingly emerging, highlights the interests of strong state actors. Although these contrasting frameworks often lead to similar interventions, the former may overly disregard politics in the pursuit of apolitical humanitarian goals whereas the latter may rely too heavily on outside political interests to guide programmatic and research decisions. Given global concerns about peace, violent extremism, and war, it is crucial that as the field moves forward, researchers continue to examine these theoretical paradigms that serve to shape as well as respond to real life events.
Footnotes
Notes
Authors
DANA BURDE is an associate professor and director of international education at New York University’s Steinhardt School of Culture, Education, and Human Development, Department of Humanities and Social Sciences; and an associate faculty member of the Politics Department; 246 Greene Street, Suite 300, New York, NY 10003, USA; email:
.
AMY KAPIT is the coordinator and lead researcher for Education Under Attack 2018, the flagship publication of the Global Coalition to Protect Education from Attack (GCPEA), which tracks violations against students, educational personnel, and educational institutions worldwide. Previously, she was the research director for a mixed-methods, randomized controlled trial studying the sustainability of community-based schools in Afghanistan, funded by a large grant from the U.S. Agency for International Development (USAID; principal investigator, Dana Burde; Co–principal investigators, Joel Middleton and Cyrus Samii). Her academic research focuses on how humanitarian organizations address attacks on education, including violence, harassment, and threats against students, teachers, and schools in areas of armed conflict Her work has appeared in publications for GCPEA and for USAID. She holds a BA in religion and peace and conflict studies from Swarthmore College and a PhD in international education from New York University’s Steinhardt School of Culture, Education, and Human Development. The author can be contacted at 350 Fifth Avenue, 34th Floor, New York, NY 10118, USA; email:
RACHEL L. WAHL is an assistant professor in the Department of Leadership, Foundations and Policy at the University of Virginia’s Curry School of Education, 405 Emmet St S, Charlottesville, VA 22904, USA; email:
OZEN GUVEN is a doctoral candidate in international education at New York University’s Steinhardt School of Culture, Education, and Human Development, Department of Humanities and Social Sciences, 246 Greene Street, Suite 300, New York, NY 10003, USA; email:
MARGOT IGLAND SKARPETEIG is the policy director in Human Rights of Norad, the Norwegian Agency for Development Cooperation, the technical arm of the Norwegian Ministry of Foreign Affairs, 7 Juniplassen 1, N-0032 Oslo, Norway; email: